Section 6.9      Action Feature Set

Action features either elicit an action or call for an action to be taken. Elicit actions include tags and notes, where a QC Tag is a call to action. There are two (2) types of QC Tags: automated and user-defined tags.

Section 6.9.1         Artificial Intelligence Quality Control Tags

Artificial Intelligence analyzes the images and tags those images that have either a) original hard copy content issues and/or b) images that have capture issues stemming from WIB™ Unit settings or operator errors. AI will create an automated tag for images that have one (1) or more Quality Control issues. These tags are used to filter records in the Quality Control Module for review by an independent Quality Control Specialist. AI QC Tags are defined as part of a Collection’s Taxonomy.

Section 6.9.1.1             AI Content Quality Control Tags

Below is a list of AI QC Content Tags and their definitions.

Name	Definition
ContentCutoffOriginal	The document is fully visible in the image frame, and the content appears to be cut off at the original document's physical edges due to the original records' printing/design/production limitations.
Creases	Content is obscured, distorted, or partially hidden in a crease.
Folds	Folded areas hide or significantly distort content.
HandwritingIllegible	Handwriting requires guessing or has multiple valid interpretations.
Highlighting	Highlighting interferes with character recognition or creates ambiguity.
LabelsCoveringInfo	Important document content is hidden beneath labels.
Smudges	Text becomes unreadable or ambiguous due to smudging.
Stains	Content is obscured, discolored, or made unreadable.
TapeHighGloss	Reflection or adhesive obscures the underlying content.
Tears	Text or important content is actually missing due to a tear.
TextTooLight	Faded text makes content difficult or impossible to read.
Underlining	Underlines merge with text characters, creating ambiguity.
Wrinkles	Text distortion makes characters unrecognizable or ambiguous.

Section 6.9.1.2             AI Scan Quality Control Tags

Below is a list of AI QC Scan Tags and their definitions.

Name	Instructions
BadLighting	Poor lighting creates unreadable or ambiguous text areas.
ContentCutoff	Important document areas are excluded from the scan.
Glare	Glare from lights or flash makes text unreadable.
HandObstruction	A hand or fingers visibly cover readable text that could be revealed by rescan.
MotionBlur	Text is blurred due to camera movement making characters unreadable.
ObjectObstruction	Objects covering VISIBLE textual content that impacts extraction and can be removed during rescan.
Overexposed	Text is washed out or invisible due to excessive brightness.
Shadows	Shadows obscure text or create difficulties reading.
TextCutoff	Important text is cut off at image edges due to poor framing and not from the original document itself.
Underexposed	The text is too dark to read clearly due to insufficient light.
UnevenLighting	Uneven lighting across the image affects text readability.

Section 6.9.2        User-defined QC Tag

The QC Tag allows the reviewer to communicate with the operator capturing the images to either reprocess or append, add to, a specific image and/or record. The QC tag sends a message to the operator. We encourage the quality review process to happen within 24 hours of the record appearing on the review platform. The QC Tag requires a message to convey the information as well as an action request. The following actions can be requested: add image, replace image. Do not use the QC tag to show the condition of a record that cannot be corrected by adding to the record or reprocessing the record. Record conditions should be added as a tag or a note to the record.

Section 6.9.2.1          Tag a Record/Image (Roadmap)

A user can add a tag to a record. Tags can be public or private. Tags allow users to mark records with a commonality. Private tags can only be viewed by the user who added the tag. A private tag can be made public if needed.

Section 6.9.3        View AI QC Tags

The tags are displayed in the Search grid, Preview Pane, and the Tags Widget as Tags and Automated Tags. Tags are user-defined, and Automated Tags are generated from the AI QC.

A user can choose to view Tags in Review in addition to the Quality Control Production Module. To view the Tags in Review, select the Tags widget from the Configure Tiles dialog. The user can also see the entire AI QC synopsis by selecting the Quality tab in the image viewer.

Configure Tiles                                                                      Tags Widget

​

AI Quality Analysis Synopsis